ARTICLES

Subjective Well-Being and Political Participation:A Comparison of Unemployed and Employed Youth

Jasmine Lorenzini

� Springer Science+Business Media Dordrecht 2014

Abstract In this paper I analyze the role of subjective well-being in unemployed and

employed youth political participation. Research shows that life satisfaction increases par-

ticipation in voting, but has no effect on protest activities when looking at the overall pop-

ulation. However, in the case of youth, life dissatisfaction fosters the potential for protest

activities. Since unemployment is detrimental for the subjective well-being of individuals,

especially when long-lasting, I ask whether the reduced subjective well-being of long-term

unemployed youth, their life dissatisfaction, fosters their participation in two forms of voice-

based participation—contacting and protest activities—that can be used to express their

dissatisfaction. I find that life dissatisfaction fosters the participation of employed youth in

contacting activities, but not that of unemployed youth. Quite on the contrary, for protest

activities, it is life satisfaction that fosters participation of the unemployed youth.

Keywords Youth � Unemployment � Life satisfaction � Political participation

1 Introduction

Youth unemployment in Europe is growing and lasting longer. Thus, unemployment can

lead to youth long-term exclusion from the labor market. Yet, unemployed youth have not

become a highly visible group on the political stage. Unemployed youth are not taking the

streets to express their dissatisfaction with the lack of opportunities for youth to enter the

labor market and neither are they turning to political action to protest against unemploy-

ment rates that can be as high as 50 % among youth in southern Europe. In this context, it

is all the more important to understand how unemployment contributes to youth political

participation.

J. Lorenzini (&)InCite - Institute of Citizenship Studies, University of Geneva, Bd du Pont-d’Arve 40, 1211 Geneva 4,Switzerlande-mail: jasmine.lorenzini@unige.ch

123

J Happiness StudDOI 10.1007/s10902-014-9514-7

Research has found that unemployment reduces life satisfaction (see Dolan et al. 2008

for a literature review) and that life dissatisfaction fosters youth potential for protest

activities (Marsh 1977; Barnes et al. 1979). However, this knowledge about the negative

impact of unemployment on subjective well-being has not been applied to the study of

unemployed youth political participation. Research in this field has found that unemployed

youth were more dissatisfied with politics (Bay and Blekesaune 2002; Bynner and Ashford

1994) and displayed more favorable attitudes toward radical political action and law-

lessness (Clark 1985; Breakwell 1986; Gaskell and Smith 1985). However, unemployed

and employed youth share similar degrees of political participation in voting (Banks and

Ullah 1987; Bynner and Ashford 1994) and the former are not more likely to vote for

extreme right or left political parties (Banks and Ullah 1987).

In this paper, I focus on voice-based political participation defined as forms of political

participation used to explicitly express an opinion or a preference (Teorell et al. 2007). More

specifically, I analyze protest and contacting activities. According to Verba et al. (1995),

protest and contacting activities permit the transmission of specific messages by groups with

fewer political resources. Through protest activities citizens transmit specific messages by

taking the streets and through contacting activities they do so by writing to elected bodies or

representatives. Thus, contacting and protest activities are forms of political participation that

unemployed youth could use to express discontent with their current situation.

I analyze the situation of long-term unemployed youth in Switzerland. Unlike Spain,

Greece, or Italy, Switzerland has a low youth unemployment rate of less than 10 %.

Nonetheless, Switzerland is an interesting case to analyze when studying the relationship

between unemployment and political participation, even more so when focusing on the role

of life satisfaction. Youth who are unemployed, in Switzerland, are confronted with a

strong social norm of work (Chabanet 2007), which reduces the unemployed persons’

subjective well-being (Stutzer and Lalive 2004). Moreover the effects of unemployment on

subjective well-being are stronger in Switzerland than in other contexts (Frey and Stutzer

2000). Thus, the Swiss context offers an interesting case for the study of the interplay

between the personal and the political consequences of youth long-term unemployment.

More specifically, I conduct this research on long-term unemployed and employed youth

living in one Swiss canton,1 namely Geneva. The political context in Geneva is representative

of the Swiss political context, characterized by political participation through direct demo-

cratic tools, moderate public demonstrations, and signing petitions (Kriesi et al. 1995; Kriesi

2008). Working, more specifically, on unemployed persons, della Porta (2008) found that

unemployed persons used both conventional and demonstrative protest forms of political

participation in Switzerland and Giugni (2008) showed that the Swiss political context

offered low to moderate opportunities for claim making related to unemployment.2 The major

implication for my study is that I test the contribution of subjective well-being in a context of

intermediary to low mobilization of unemployed persons. Therefore, if any effects of sub-

jective well-being are evidenced in a context of low mobilization, it can be expected that these

effects will be stronger in contexts of higher mobilization.

This research contributes to the literature on subjective well-being and political par-

ticipation by showing the effects of life satisfaction on specific forms of political partic-

ipation enacted by certain social groups. Contrary to Flavin and Keane (2012) who find that

life satisfaction does not contribute to protest activities, I find that life satisfaction fosters

1 In the Swiss federal state, the term ‘‘canton’’ denotes the member states.2 Claim making is defined as ‘‘the expression of a political opinion by verbal or physical action in the publicspace’’ (Giugni 2008: 302).

J. Lorenzini

123

unemployed youth protest activities. In addition, I find that life satisfaction hinders

employed youth contacting activities. The comparison of unemployed and employed youth

shows that the effects of subjective well-being on political participation vary across social

groups. Lastly, I test four causal mechanisms to explore the effects of subjective well-being

on political participation.

2 Theoretical Framework

2.1 Subjective Well-Being and Political Participation

A wide array of studies in economy and psychology has addressed life satisfaction. In this

literature, the authors have referred to life satisfaction and happiness as equivalent terms

that can be defined as ‘‘[denoting] the degree to which people judge the overall quality of

their life as a whole favourably’’ (Veenhoven 1988: 334). Moreover, Frey and Stutzer

considered life satisfaction the best empirical approximation to the concept of subjective

well-being (2010). These studies analyzed the role of social and personal factors in

explaining life satisfaction (see Dolan et al. 2008 for a literature review).

Although less research has been conducted on the links between life satisfaction and

politics, scholars working on political participation have shown a growing interest for

subjective well-being potential to explain democratic life. Several researchers have used

subjective well-being to explain political participation (Flavin and Keane 2012; Weitz-

Shapiro and Winters 2011; Barnes et al. 1979). Others have analyzed it as resulting from

political institutions and participation (Tavits 2008; Frey and Stutzer 2000; Dorn et al.

2008). I discuss both approaches and their contribution to the study of life satisfaction and

political participation.

In a precursory work on happiness, Veenhoven (1988) explained that happy people were

more satisfied with various aspects of their lives and in turn were more empathetic and

socially sensitive; they were less absorbed by personal problems. He then turned to the

consequences of happiness for democracy and discussed two potential outcomes. On one

hand, he discussed the idea that happy people withdrew from politics and that discontent

drove political participation. On the other hand, Veenhoven (1988) proposed that happy

people were more involved in the community, they felt more concerned by social and

political problems, and were less preoccupied with themselves.

Research has found empirical support for both arguments. Indeed, the effects of life

satisfaction vary across forms of political participation. Flavin and Keane (2012) found that

life satisfaction increased individual political participation with regard to voting and insti-

tutional forms of participation, but did not contribute (positively or negatively) to protest

activities. Yet Barnes et al. (1979) found that the influence of personal dissatisfaction on

protest varied across sub-groups of the population. Personal dissatisfaction, measured

through dissatisfaction with one’s life as a whole, contributed to youth protest potential.

Barnes et al. (1979) did not work on behaviors, they worked on protest potential measured

through approval of seven protest activities, including signing petitions, joining boycotts,

attending demonstrations, refusing to pay taxes or rent, joining wildcat strikes, occupying

buildings, and blocking traffic. Moreover, Marsh (1977) confirmed that the effects of life

dissatisfaction on protest potential were significant for certain age groups. He added that

variations also appeared among social classes and political affiliations: middle- and working-

class supporters of the left were more keen to protest when they were dissatisfied. Thus to

understand the effects of life satisfaction on forms of political participation other than voting,

Subjective Well-Being and Political Participation

123

one needs to consider that life satisfaction may only influence the participation of certain

groups of the population. Here I propose to test this argument analyzing specific political

behaviors of long-term unemployed and employed youth. I will examine protest activities,

focusing on taking part in public demonstrations, as well as on contacting activities.

However, some works have contended that life satisfaction derived from political

participation and the specificities of political institutions. Tavits (2008) showed that in

European countries, being represented by candidates for whom one voted and living in a

country with low corruption rate increased individual happiness. Frey and Stutzer (2000)

studied the specificities of the Swiss direct democracy and showed that in cantons offering

more direct democratic rights, people tended to be happier. Yet, Dorn et al. (2008) showed

that these findings were no longer statistically significant when controlling for cultural

differences across the country.3 Weitz-Shapiro and Winters (2011) found empirical support

for the argument that life satisfaction affected political participation. They used three

alternative methods to test the direction of the link between voting and life satisfaction in

Latin America and found that life satisfaction caused political participation. Additionally,

Thoits and Hewitt (2001) analyzed engagement in volunteering depending on life satis-

faction. They found that life satisfaction supported civic engagement. In this study, Thoits

and Hewitt (2001) also discussed the direction of the causality and tested two alternative

hypotheses: that subjective well-being contributed to civic engagement, or that civic

engagement increased subjective well-being. They found support for a model of causality

running from subjective well-being to political participation.

Research on unemployed youth political participation has not analyzed their subjective

well-being. These studies focused on unemployed youth political attitudes (Clark 1985;

Breakwell 1986; Gaskell and Smith 1985) and their voting behaviors (Bynner and Ashford

1994; Banks and Ullah 1987) to understand whether unemployed youth were more dis-

satisfied with politics, less politically active than employed youth, or supportive of extreme

political parties. Hence, I address three research questions related to subjective well-being

and political participation: Does life dissatisfaction foster youth voice-based political

participation? Is the effect of life dissatisfaction stronger for unemployed youth than for

employed youth? How does life dissatisfaction contribute to political participation?

2.2 Subjective Well-Being and Unemployment

Since the seminal study on the consequences of unemployment for the inhabitants of

Marienthal conducted by Jahoda et al. (2002 [1932]) many studies have reassessed the

detrimental consequences of unemployment for subjective well-being (Clark and Oswald

1994; Winefield 1997; Whelan and McGinnity 2000; Oesch and Lipps 2012; Ervasti and

Venetoklis 2010). Moreover Winkelmann and Winkelmann (1998) have convincingly

demonstrated, with the use of panel data, that the direction of causality ran from unem-

ployment to reduced life satisfaction. Their study ruled out the self-selection hypothesis

stating that unemployed persons became unemployed due to their subjective well-being, as

well as the omitted variable hypothesis proposing that third variables, health-related or

family-related problems, contributed to both unemployment and subjective well-being. In

their literature review of factors associated with life satisfaction, Dolan et al. (2008)

3 Running the same analyses as Frey and Stutzer (2000) while taking into account the Swiss multilingualdivision, indicators of cantonal and individual language, as well as controls for religion and income, Dornet al. (2008) found that the effects of direct democracy institutions were no longer significant predictors ofhappiness.

J. Lorenzini

123

discussed other studies that came to the same conclusion. Therefore, it is widely accepted

that unemployment reduces life satisfaction.

3 Hypotheses

3.1 Testing the Effect of Subjective Well-Being on Political Participation

My hypotheses are derived from studies that used life satisfaction as an independent variable.

According to this literature, two expectations can be formulated. On the one hand, life

satisfaction may foster participation in voting (Flavin and Keane 2012; Weitz-Shapiro and

Winters 2011). On the other hand, life satisfaction may hinder participation in protest. In other

words, it is life dissatisfaction that fosters protest activities (Barnes et al. 1979; Marsh 1977).

However, a qualification has to be made regarding the latter argument, since life dissatis-

faction has been found to have an effect only on the protest potential of specific groups, youth,

working and middle class, but not for the general population. Indeed, working on the overall

population, Flavin and Keane (2012) found no effect of life satisfaction on protest activities.

Here, I analyze the political participation of unemployed and employed youth in two

alternative voice-based forms of participation; working on two forms reveals whether the

results are contingent upon a specific form of political participation. Working on a specific

population for which life dissatisfaction increases the potential for protest, youth, I expect

that life dissatisfaction fosters youth’s participation in both voice-based forms of partici-

pation. My first hypothesis is that youth who are dissatisfied with their lives participate

more in protest and contacting activities.

3.2 Comparing the Effect of Subjective Well-Being on Political Participation Across

Employment Statuses

Life dissatisfaction increases youth potential for protest, but also that of individuals per-

taining to certain social classes: the working and middle class. This can be related to

variations in the relative deprivation faced by individuals pertaining to different social

classes. Subsequently, I anticipate the effect of life dissatisfaction to be stronger for

unemployed youth than for employed youth. My second hypothesis states that the effects

of life satisfaction are stronger for the unemployed youth than for the employed youth.

3.3 Exploring Causal Mechanisms

I explore four alternative causal mechanisms linking life satisfaction to political partici-

pation. I test the effects of life satisfaction on political participation when controlling for

socioeconomic status, civic attitudes, social capital, and psychological well-being. The

effects of omitted predictors may be captured by life satisfaction when they are not

included in the model. So including them may reveal causal mechanisms through which

life satisfaction affects political participation.

Firstly, life satisfaction can vary across socioeconomic groups; for instance those who

face less financial difficulties may be less dissatisfied with their lives (Boes and Winkel-

mann 2010). Moreover, in examining the effects of unemployment on political partici-

pation, it is important to control for socioeconomic status. Burden and Wichowsky (2012)

found that the unemployed were less likely to vote in presidential elections in the US than

employed individuals. Yet, in their views this is due to their socioeconomic status and not

Subjective Well-Being and Political Participation

123

to the experience of unemployment. Indeed, Scott and Acock (1979) showed that the

impact of unemployment on political participation varied depending on the socioeconomic

status of the unemployed. Schlozman and Verba (1979), analyzing different forms of

political participation, concluded that unemployment was more prevalent among certain

socioeconomic groups. In their view, what affected political participation was not the

unemployment, but other characteristics of the persons confronted with unemployment.

Hence, including predictors of socioeconomic status allows me to test the first causal

mechanism—life satisfaction affects political participation due to its correlation with

socioeconomic statuses—and to test the effects of life satisfaction while controlling for an

important set of predictors of political participation.

Secondly, life satisfaction varies depending on the satisfaction with political repre-

sentation. Tavits (2008) found that citizens were more satisfied with their lives when they

voted for the elected representatives and live in a context of low corruption. Furthermore,

research in political science has found that civic attitudes affect political participation

(Almond and Verba 1963; Dalton 2006; Conway 2000). These arguments are related to the

socioeconomic model, which is the dominant model in the field of political participation

(Leighley 1995). Socioeconomic status contributes to both political resources and civic

attitudes, which in turn spur political participation (Verba et al. 1995). Civic attitudes are

‘‘attitudes which individuals hold toward themselves or the political system which pre-

dispose them toward political action’’ (Leighley 1995: 183). Including civic attitudes

permits to test the second causal mechanism according to which life satisfaction captures

the effects of civic attitudes. Moreover, it permits to test the effect of life satisfaction while

controlling for a second set of important predictors of political participation.

Thirdly, the literature working on the consequences of unemployment has shown that

unemployment hindered social inclusion (Gallie et al. 2003; Paugam 2006; Hammer 2000) and

research on life satisfaction has shown that social inclusion contributed to life satisfaction

(Dolan et al. 2008). Regarding protest activities, analyses of collective mobilizations by the

unemployed have shown the importance of organizational resources derived from associational

membership (Maurer 2001; Chabanet 2008; Faniel 2004). Therefore, in accounting for the role

of subjective well-being for unemployed youth political participation, I include in my analysis

measures of involvement in civil society organizations. Thus, it allows testing the third causal

mechanism, life satisfaction fosters political participation due to the association of life satis-

faction with social capital. Again including these predictors permits to test the effect of life

satisfaction on political participation while controlling for predictors of political participation.

Fourthly, life satisfaction is related to psychological well-being (Dolan et al. 2008) and

psychological well-being is also expected to account for unemployed adult political par-

ticipation (Rosenstone 1982; Rosenstone and Hansen 1993). Analyzing the 1974 presi-

dential election in the US, Rosenstone (1982) found that unemployed persons were less

likely to vote. Although the effect was stronger during the first phase of unemployment

when individuals needed to register and invest time in the administrative work related to

unemployment, the lower propensity to vote lasted throughout the unemployment span. He

contended that the unemployed persons’ main concerns were related to solving their

personal problems and highlighted the psychological impact of unemployment to explain

their lower propensity to vote. In a later study, Rosenstone and Hansen (1993) found that

unemployed persons were less likely to engage in contacting activities and used the same

argument to explain their lack of contacting. I include a measure of mental health based on

the General Health Questionnaire that can be used to assess psychological well-being. This

allows me to test the fourth causal mechanism that is life satisfaction captures the effect of

psychological well-being on political participation.

J. Lorenzini

123

4 Method

4.1 Procedure and Participants

I use original survey data on long-term unemployed youth residing in Geneva constructed

within a European research project on youth unemployment and exclusion (YOUNEX).

The data resulted from computer-assisted telephone interviews conducted on 301 long-

term unemployed youth and 317 employed youth residing in Geneva at the time of the

interview in 2010. Reaching long-term unemployed youth in a context of a low youth

unemployment is not an easy task. The survey included two sub-samples of unemployed

youth. The first sub-sample was drawn from the population of Geneva residents aged

18–34 years old; long-term unemployed youth were found through a screening procedure

including questions on their employment status. The second sub-sample was drawn from a

list of long-term unemployed youth constructed by contacting the registered unemployed

youth through a mailing and recruiting them in the unemployment office. I tested the two

sub-samples with regard to key sociodemographic characteristics, social and political

inclusion measures, as well as subjective well-being, to confirm that I could merge both

samples.4 The control group of employed youth included youth aged 18–34 who had an

open-ended contract for at least one year. The sample of employed youth was drawn from

the youth population of Geneva and was constructed through screening as well. The survey

included questions on employment status, social and political inclusion, subjective well-

being, and sociodemographic information.

4.2 Measures

4.2.1 Dependent Variables

Protest and contacting activities are dichotomous variables. I measure protest activities

through a single item: taking part in public demonstrations.5 I code protest as 1 when the

respondents participated in public demonstrations. I code the respondent’s contact score as

a 1 when the respondent engaged in any one or more of the following political acts:

contacting a politician, a national or local government official, the media, and/or a judicial

body (for non personal reasons). The Cronbach’s alpha for the contacting activities index

reaches conventional levels (.615). Descriptive statistics for unemployed and employed

youth on all the dependent and independent variables are presented in Appendix 1.

4.2.2 Independent Variable

The main independent variable measures subjective well-being. The survey measures

subjective well-being with a question on life satisfaction. The question asked, ‘‘Taking all

things together, how happy would you say you are?’’ on a scale from 0 to 10. Following

Dorn et al. (2008: 233), I aggregate the four lowest categories (from 0 to 3) to avoid

4 Additional information about the sample design and the tests I conducted to merge the sub-samples oflong-term unemployed youth can be found in my doctoral thesis (Lorenzini 2013).5 For protest activities, I also tried to construct an index using participation in illegal and violent actions inaddition to participation in public demonstration, but the Cronbach’s alpha for this index was rather low(.385). Therefore, I decided to focus only on participation in public demonstration the most widespreadprotest activity among the three.

Subjective Well-Being and Political Participation

123

inference problems due to the small number of individuals in these categories. Hence, I

measure life satisfaction on a scale from 0 to 7, where 0 reflects life dissatisfaction and 7

the highest degree of life satisfaction. Life satisfaction is normalized on a 0–1 scale for the

analyses.

4.2.3 Control Variables

The respondents’ socioeconomic statuses include sex, age, nationality, education, and

financial difficulties. Sex is a dichotomous variable coded 0 for male and 1 for female. Age

ranges from 18 to 35 years old, but is normalized on a 0–1 scale. Nationality is a

dichotomous variable coded 0 for foreigners and 1 for Swiss citizens. I measure education

through a three-state categorical variable measuring educational levels. The three states are

below secondary education (the reference category), secondary education, and tertiary

education. I measure financial difficulties through a dichotomous variable based on a

subjective question asking to what extent it is difficult to cope on present income.

Civic attitudes include political interest, political efficacy (internal and external), and

left–right self-placement. I measure political interest through a subjective question asking

the respondent to what extent he or she was interested in politics. Two dichotomous

variables measure the political efficacy: internal political efficacy measuring one’s per-

ceived influence on politics and external political efficacy addressing political actors’

responsiveness. The three dichotomous variables are coded 0 when the political attitude is

absent and 1 when the respondent is either interested in politics or feels he or she has

political efficacy. I measure left–right self-placement through a three-state categorical

variable. I compare those individuals who placed themselves on the left or on the right of

the scale to those who did not place themselves on the scale or chose the middle point,

which I use as reference category.

I measure social inclusion through household composition, interpersonal networks, and

associational networks. The household composition includes two dichotomous variables:

living with one’s partner and living with children. A three-state categorical variable on

frequency of contacts with friends during the last month measures interpersonal networks.

The three states are: no contacts (reference category), few contacts (once or twice), and

frequent contacts (weekly or daily contacts). A dichotomous variable on membership in

civil society organizations measures associational networks.

Lastly, I measure psychological well-being through four items from the 12-item General

Health Questionnaire (GHQ). The four items correspond to one dimension identified with a

factor analysis. These items are: ‘‘I have lost much sleep over worry,’’ ‘‘I feel that I cannot

overcome my difficulties,’’ ‘‘I have been losing confidence in myself,’’ and ‘‘I think of

myself as a worthless person.’’ The index I construct with these four items has a Cron-

bach’s alpha of 0.78. I reverse the item coding so that on this additive scale, a high score

measures mental health.

4.3 Data Analyses

I use logistic regressions to test the two hypotheses: (1) life dissatisfaction fosters political

participation and (2) the effect of life dissatisfaction in fostering political participation is

stronger for the unemployed youth than for the employed youth. I construct the logistic

regression model through a stepwise procedure. The stepwise procedure allows me to

explore four causal mechanisms that could explain the effect of life dissatisfaction on

political participation. The first Model includes only employment status and life

J. Lorenzini

123

satisfaction as predictors. Then, in the second Model, I introduce an interaction effect

between employment status and life satisfaction to account for the differential effect across

groups. In the further four models, Models 3–6, I subsequently add four sets of predictors

to explore the proposed causal mechanisms related to socioeconomic status, civic attitudes,

social capital, and psychological well-being.

I test the two hypotheses using Model 7, which is the full model including employment

status, life satisfaction, an interaction term between employment status and life satisfac-

tion, as well as the socioeconomic, the civic attitudes, and the social capital predictors. I

present predictive margins to discuss the results of these analyses. For logistic regressions,

predictive margins show the changes in the probability of participating in either protest or

contacting activities for individuals who have specific characteristics (Cameron and

Trivedi 2009). I calculate both adjusted predictions and marginal effects to account for the

contribution of life satisfaction in explaining unemployed and employed youth participa-

tion in protest and contacting activities. I use predictive margins and present graphical

visualization of the effects of life satisfaction on political participation for both groups to

show the different effects across groups and forms of political participation. I use Marginal

Effects at Representative value (MER) to see the effect of moving from the lowest to the

highest value on one variable of interest for the two groups, unemployed and employed

youth, while holding all the other independent variables to the observed values.

I run multicollinearity tests on my models and find that VIF and tolerance scores show

no sign of multicollinearity. Additionally, the results based on logistic regressions are

robust, I find similar results using linear probability models following the procedure

proposed by Mood (2010), as well as full interactions (separate models for unemployed

and employed youth).

5 Results

5.1 Effects of Life Satisfaction on Political Participation

In this section, I report statistical tests addressing hypothesis 1—life dissatisfaction fosters

political participation—and hypothesis 2 anticipating a differential effect of life dissatis-

faction on the political participation of unemployed and employed youth. Figures 1 and 2

present graphically the probability that both unemployed and employed youth participate

in protest and contacting activities, while Tables 1 and 2 present MER.

In Fig. 1 we see that the effects of life satisfaction on the probability of participating in

protest activities differ across employment status groups. Life satisfaction increases the

probability that unemployed youth participate in protest activities, while for employed

youth life satisfaction reduces the probability of doing so. Yet, the difference between

unemployed and employed youth is only statistically significant at the highest levels of life

satisfaction. Among those who are satisfied with their lives, moving from employed to

unemployed significantly increases the likelihood of participating in protest activities. At

the lowest levels of life satisfaction, the difference does not reach conventional levels of

statistical significance and, among employed youth, the dispersion is large due to the small

number of employed respondents confronted with very low levels of life satisfaction.

Table 1 shows the substantial effect of life satisfaction in predicting employed youth

contacting activities. Moving from the lowest to the highest level of life satisfaction

increases unemployed youth probability of protest activities by 14.2 percentage points, and

this increase is significant at a 10 % threshold, which is acceptable for a small sample.

Subjective Well-Being and Political Participation

123

Fig. 1 Probability of doing protest activities by employment status. Source: Appendix 2, model 7 (fullmodel)

Fig. 2 Probability of doing contacting activities by employment status. Source: Appendix 3, model 7 (fullmodel)

J. Lorenzini

123

Moreover, life satisfaction reduces employed youth probabilities of protest activities by

12.4 percentage points. This reduction is not statistically significant, however the confi-

dence intervals range from -0.31 to 0.06 and the overlap of the zero is slight. Furthermore,

comparing the substantial effect of life satisfaction with other predictors shows that the

effect of life satisfaction on the probability of protest activities is equivalent, in magnitude,

to that of associational membership or left self-placement and stronger than political

interest or education. Thus, it appears that life satisfaction is among the strongest predictors

of protest activities.

For protest activities, I fail to confirm the first hypothesis. Life satisfaction fosters

protest activities for unemployed youth, while it hinders protest activities for employed

youth. Nonetheless, the analyses reveal a substantial effect of life satisfaction and confirm

the second hypothesis. I find different effects across groups. Yet, the effects do not differ in

the anticipated way. The results show that life satisfaction plays an altogether different role

for the protest activities of both groups—increasing that of unemployed youth, while

reducing that of employed youth.

Turning to Fig. 2, we see that life satisfaction also reduces employed youth participation

in contacting activities. Those employed youth who are most dissatisfied with their lives

have a higher probability of participating in contacting activities than those who are most

Table 1 Marginal effect of main predictors on protest activities by unemployed and employed youth (95 %confidence intervals in brackets)

Protest activities

Model 7

Employeddy/dx

Unemployeddy/dx

Life satisfaction -0.124[-0.31,0.06]

0.142?

[-0.02,0.30]

Socioeconomic status

Education (Ref. below secondary)

Secondary education 0.068*[0.00,0.13]

0.075*[0.00,0.15]

Tertiary education 0.116*[0.02,0.21]

0.127*[0.02,0.23]

Civic attitudes

Political interest 0.097**[0.04,0.16]

0.106***[0.04,0.17]

Left–right (Ref. center)

Left self placement 0.143***[0.07,0.22]

0.157***[0.07,0.24]

Right self placement -0.041[-0.10,0.02]

-0.046[-0.11,0.02]

Social capital

Associational membership 0.122***[0.06,0.18]

0.133***[0.07,0.19]

Source: Appendix 2, model 7 (full model)

Model 7 includes: life satisfaction, employment status, interaction term between life satisfaction andemployment status, socioeconomic status, political attitudes, and social capital? p \ 0.10; * p \ 0.05; ** p \ 0.01; *** p \ 0.001

Subjective Well-Being and Political Participation

123

satisfied with their lives. Regarding contacting activities, employed youth differ again from

unemployed youth since life satisfaction does not contribute to unemployed youth par-

ticipation in contacting activities. Unemployed youth who are most dissatisfied with their

lives have a statistically significant lower probability than employed youth to participate in

contacting activities and the reverse is true for unemployed youth who were most satisfied

with their lives.

MER presented in Table 2 confirm, on a second form of political participation, the

substantial contribution of life satisfaction. For employed youth, moving from the least to

the most satisfied with their lives reduces the probability of participating in contacting

activities by 22.4 percentage points. The marginal effect of life satisfaction for the

employed youth is higher than that of associational membership and political interest, two

important predictors of political participation. However, for the unemployed youth life

satisfaction does not contribute at all to their contacting activities (0.006).

In this second form of political participation, I fail again to confirm the first hypothesis.

Life satisfaction hinders employed youth participation in contacting activities, but it has no

effect on that of unemployed youth. Thus, it confirms the second hypothesis; life satis-

faction has a differential effect across groups. Furthermore, analyzing the effects of life

satisfaction on a second form of political participation corroborates the finding that life

Table 2 Marginal effect of main predictors on contacting activities by unemployed and employed youth(95 % confidence intervals in brackets)

Contacting activities

Model 7

Employeddy/dx

Unemployeddy/dx

Life satisfaction -0.224*[-0.40, -0.04]

0.006[-0.14, 0.16]

Socioeconomic status

Education (Ref. below secondary)

Secondary education 0.054[-0.02, 0.13]

0.058[-0.02, 0.14]

Tertiary education 0.045[-0.04, 0.13]

0.049[-0.05, 0.15]

Civic attitudes 0.074*

[0.01, 0.13]0.081*

[0.02, 0.14]

Political interest

Left–right (Ref. center)

Left self placement 0.034[-0.03, 0.10]

0.037[-0.04, 0.11]

Right self placement 0.080*[0.01, 0.15]

0.087*[0.00, 0.17]

Social capital

Associational membership 0.116***[0.06, 0.17]

0.125***[0.06, 0.19]

Source: Appendix 3, model 7 (full model)

Model 7 includes: life satisfaction, employment status, interaction term between life satisfaction andemployment status, socioeconomic status, political attitudes, and social capital? p \ 0.10; * p \ 0.05; ** p \ 0.01; *** p \ 0.001

J. Lorenzini

123

satisfaction has a substantial effect more important than political interest, left–right self-

placement, and associational membership.

5.2 Exploring Causal Mechanisms

Turning to the exploration of causal mechanisms, Tables 3 and 4 present the effects of

moving from the lowest to the highest level of life satisfaction for both unemployed and

employed youth on the probability of participating, respectively, in protest and contacting

activities with different sets of control variables.

MER presented in Table 3 are calculated from the seven logistic regressions models

analyzing protest activities.6 In Table 3, we see that the effects of life satisfaction on

unemployed youth protest activities only appear in the full model controlling for socio-

economic status, civic attitudes, and social capital. In Models 3–5, MER fail to reach

statistical significance. However, the confidence intervals seem to point at a positive effect

of life satisfaction on unemployed youth protest activities, life satisfaction appears to

increase unemployed youth participation in protest activities. For instance, in Model 3, the

confidence intervals range from -0.06 to 0.29. Thus, the confidence intervals suggest that

the effect is likely to be positive for unemployed youth and this effect may become

statistically significant with a bigger sample.

Moving to the specific contribution of different sets of predictors to protest activities,

we see that rather than reducing the effects of life satisfaction they permit to reveal them

(Model 7). First, in Model 2 with the interaction effect between life satisfaction (the

variable of interest) and employment status (the moderator), we see that the effect of life

satisfaction is likely to differ across groups. While the coefficients are not statistically

significant, the confidence intervals reveal divergent tendencies—life satisfaction reduces

the protest activities of employed youth (with CI ranging from -0.28 to 0.10) and

increases those of unemployed youth (with CI ranging from -0.09 to 0.23). This

divergence persists in further models including different sets of predictors, Models 3–6,

and the coefficient becomes statistically significant for the unemployed youth in the last

model.

Adding socioeconomic status, civic attitudes, and social capital discloses the effect of

life satisfaction on protest activities. Thus, life satisfaction is not capturing the effect of

other sets of predictors and cannot be related to the four causal mechanisms proposed.

Substantially, we see that civic attitudes contribute the most to the model with a pseudo

r-square around 0.119 for Model 4. Socioeconomic status and social capital also contribute

to the model, although in smaller proportions with pseudo r-squares of, respectively 0.041

for Model 3 and 0.066 for Model 5, while in Model 6 we see that psychological well-being

adds little to the model (0.009). Comparing the fit of the different models using the Akaike

Information Criterion (AIC), the lower the AIC, the better the model fit, confirms that the

predictors added to Model 3, 4, and 5 improve the model. But psychological well-being

does not improve the model, in Model 6 the AIC is bigger. Lastly, Model 7 has the lowest

AIC and a pseudo r-square of 0.187.

In Table 4, the MER calculated for the 7 Models analyzing contacting activities show

that adding the different sets of predictors does not reduce the substantial effect of life

satisfaction.7 Model 1, without the interaction between life satisfaction and employment

status, shows a significant effect of life satisfaction for both groups. Life satisfaction

6 Appendix 2 presents the seven Models of the logistic regressions.7 Appendix 3 presents the seven Models of the logistic regressions.

Subjective Well-Being and Political Participation

123

reduces the contacting activities of both unemployed and employed youth and is significant

at a 10 % level. Yet, introducing the interaction between both variables in Model 2 reveals

that the effect of life satisfaction differs across groups. Life satisfaction reduces the

probability that employed youth participate in contacting activities by 24.3 percentage

points and is significant at a 5 % level, while it has no significant effect for unemployed

youth. In Model 3–5, the different effects across groups are confirmed and, importantly, the

effect of life satisfaction on the probability that employed youth participate in contacting

activities does not change, it remains over 20 percentage points. Apart from the change

between Model 1 and Model 2, when I introduce the interaction term, we see that changes

in the marginal effects of life satisfaction are minimal. Thus, the stepwise analysis shows

that the effect of life satisfaction is not capturing that of socioeconomic status, civic

attitudes, social capital, or psychological well-being. Thus, this exploratory analysis of

causal paths does not permit to uncover mechanisms through which life satisfaction

contributes to political participation.

Lastly, regarding the fit of the model, it improved slightly when adding the interaction

effect and is further improved with the set of predictors included in Models 3–5, but not in

Model 6 testing the effect of psychological well-being. Again Model 7 has the lowest AIC,

thus, it offers the best model fit. Moving to the contribution of different sets of predictors to

the models, we see that socioeconomic status, civic attitudes, and social capital contribute

similarly to the models with pseudo r-square around 0.07, while psychological well-being

adds little to the model. Model 7, the full model, has a pseudo r-square of 0.162.

Table 3 Marginal effect of life satisfaction on protest activities by unemployed and employed youth (95 %confidence intervals in brackets)

Protest activities

Model 1dy/dx

Model 2dy/dx

Model 3dy/dx

Model 4dy/dx

Model 5dy/dx

Model 6dy/dx

Model 7dy/dx

Life satisfaction

Employed 0.008[-0.10,

0.12]

-0.089[-0.28,

0.10]

-0.103[-0.29,

0.08]

-0.114[-0.30,

0.07]

-0.089[-0.28,

0.10]

-0.125[-0.31,

0.07]

-0.124[-0.31,

0.06]

Unemployed 0.009[-0.12,

0.13]

0.071[-0.09,

0.23]

0.114[-0.06,

0.29]

0.081[-0.07,

0.23]

0.094[-0.07,

0.26]

0.032[-0.14,

0.21]

0.142?

[-0.02,0.30]

AIC 489.0 489.3 484.0 444.1 469.6 489.1 433.0

Pseudo R2 0.001 0.005 0.041 0.119 0.066 0.009 0.187

N 570 570 570 570 570 570 570

Source: Appendix 2, model 1–7 with the following independent variables

M1: life satisfaction and employment status

M2: life satisfaction, employment status, and interaction term between the two variables (=main predictors)

M3: main predictors ? socioeconomic status (education, financial difficulties, sex, age, nationality)

M4: main predictors ? civic attitudes (political interest, left–right self placement, internal and externalpolitical efficacy)

M5: main predictors ? social capital (partner, children, contacts with friends, associational membership)

M6: main predictors ? mental health (GHQ)

M7: main predictors ? socioeconomic status, political attitudes, and social capital? p \ 0.10; * p \ 0.05; ** p \ 0.01; *** p \ 0.001

J. Lorenzini

123

6 Discussion and Conclusion

In concluding this paper, I discuss the implications of my findings for research on life

satisfaction and political participation, but also to understand more generally why young

unemployed in Europe are not a highly visible group. Young citizens are confronted with

rising unemployment rates across European countries. Thus, as I argued in the introduc-

tion, it is important to understand why unemployed youth are not expressing political

dissatisfaction stemming from their situation on the labor market by taking to the streets. In

this paper, I have dealt with the specific role played by life satisfaction in explaining

unemployed and employed youth voice-based political participation. More specifically, I

have analyzed the effects of life satisfaction on protest and contacting activities, two forms

of voice-based participation. Thus, I addressed the question of why unemployed youth are

not protesting by focusing on the mediating role of life satisfaction.

My analysis shows the contribution of subjective well-being to youth political partic-

ipation, while controlling for predictors that influence life satisfaction (Dolan et al. 2008)

and political participation (Verba et al. 1995). Contrary to the results found by Flavin and

Keane (2012), it appears that life satisfaction contributes both to protest activities and to

contacting activities. However, life satisfaction contributes to voice-based political par-

ticipation in a more complex way than expected.

Following the existing literature on personal dissatisfaction and protest (Marsh 1977; Barnes

et al. 1979), my first hypothesis states that life dissatisfaction fosters voice-based political

Table 4 Marginal effect of life satisfaction on contacting activities by unemployed and employed youth(95 % confidence intervals in brackets)

Contacting activities

Model 1dy/dx

Model 2dy/dx

Model 3dy/dx

Model 4dy/dx

Model 5dy/dx

Model 6dy/dx

Model 7dy/dx

Life satisfaction

Employed -0.099?

[-0.21,0.01]

-0.243*[-0.43,

-0.05]

-0.232*[-0.42,

-0.05]

-0.243**[-0.43,

-0.06]

-0.243**[-0.43,

-0.06]

-0.266**[-0.46,

-0.07]

-0.224*[-0.40,

-0.04]

Unemployed -0.107?

[-0.22,0.00]

-0.020[-0.17,

0.13]

0.043[-0.12,

0.20]

-0.030[-0.18,

0.12]

-0.017[-0.16,

0.13]

-0.050[-0.21,

0.11]

0.006[-0.14,

0.16]

AIC 478.3 476.6 462.3 459.2 458.0 477.5 439.5

Pseudo R2 0.009 0.017 0.072 0.074 0.077 0.019 0.162

N 570 570 570 570 570 570 570

Source: Appendix 1, model 1–7 with the following independent variables

M1: life satisfaction and employment status

M2: life satisfaction, employment status, and interaction term between the two variables (=main predictors)

M3: main predictors ? socioeconomic status (education, financial difficulties, sex, age, nationality)

M4: main predictors ? civic attitudes (political interest, left–right self placement, internal and externalpolitical efficacy)

M5: main predictors ? social capital (partner, children, contacts with friends, associational membership)

M6: main predictors ? mental health (GHQ)

M7: main predictors ? socioeconomic status, political attitudes, and social capital

* p \ 0.05; ** p \ 0.01; *** p \ 0.001

Subjective Well-Being and Political Participation

123

participation. Partly confirming this hypothesis, I found that life dissatisfaction fosters employed

youth participation in contacting activities, but not in protest activities. Yet, for unemployed

youth, life dissatisfaction does not foster either protest or contacting activities. Quite on the

contrary, life dissatisfaction hinders participation in protest, falsifying my first hypothesis.

My results reveal that the more satisfied with their lives, the more the unemployed youth

engage in protest activities. Overall few long-term unemployed youth engage in protest

activities: on average 17.2 % of them have done protest activities in the last twelve months.

Nonetheless, I find that life satisfaction fosters their participation in protest activities. More-

over, the effect of life satisfaction on protest activities is comparable, in magnitude, to that of

associational membership, slightly lower than that of left–right self-placement, and higher than

that of political interest. Additionally, life satisfaction appears as an even more important

predictor of contacting activities for employed youth. Life satisfaction reduces the probability

that employed youth participate in contacting by 22.4 percentage points. This change is

important considering that the average rate of contacting activities by employed youth is of

15.4%. In fact, this effect is stronger than that of political interest, left–right self-placement, and

associational membership, which are important predictors of political participation (Verba et al.

1995). Thus, my analysis shows that life satisfaction is important to consider when studying the

political participation of specific social groups. Furthermore, comparing two groups of

youngsters, long-term unemployed youth and employed youth, my analyses confirm that life

satisfaction only affects the voice-based political participation of certain groups (Barnes et al.

1979; Marsh 1977). Further analyses on other social groups are needed to extend these findings.

Although these results are interesting, they are also challenging. In this paper, I explored

four causal mechanisms to understand how life satisfaction contributes to youth political

participation. However, none of the proposed explanations find empirical support. First, I

controlled for socioeconomic status to account for variations in terms of life satisfaction

across socioeconomic statuses (Boes and Winkelmann 2010) and to control for socioeco-

nomic status effects on political participation (Verba et al. 1995). However, the effect of life

satisfaction on the protest activities of unemployed youth does not change when controlling

for socioeconomic status. In the next model, I control for civic attitudes associated with life

satisfaction (Tavits 2008) and political participation (Verba et al. 1995; Leighley 1995). Yet

again including civic attitudes does not reduce the effect of life satisfaction for the protest

activities of unemployed youth. Then, I included controls to account for social inclusion, both

in associational and interpersonal networks, which increase life satisfaction (Dolan et al.

2008) and, according to the social capital literature, political participation (Chabanet 2008;

Faniel 2004). But including predictors associated with social capital did not reduce the effect

of life satisfaction on protest activities by unemployed youth. Lastly, I controlled for psy-

chological well-being, which is expected to have an effect on life satisfaction (Dolan et al.

2008) and on the political participation of unemployed persons (Rosenstone 1982; Rosen-

stone and Hansen 1993). The exploration of this last causal mechanism shows again that the

effect of life satisfaction in predicting unemployed youth protest activities remains.

Regarding contacting activities by employed youth, the causal mechanisms explored reveal

the effects of life satisfaction. Hence, life satisfaction captures a specific mechanism con-

tributing in explaining contacting activities. But this effect only appears when differentiating

social groups and controlling for socioeconomic, civic attitudes, and social capital jointly.

Thus, my analysis of both protest and contacting activities shows that the effects of life

satisfaction on political participation are not due to socioeconomic resources, to civic

attitudes, to social capital, or to psychological well-being. This raises an important ques-

tion: Why does life satisfaction foster unemployed youth protest activities? More research

is needed to understand how life satisfaction contributes to the political participation of

J. Lorenzini

123

unemployed youth, but also to that of employed youth. Future studies should explore the

possibility that, in the case of protest activities, political participation fosters life satis-

faction. In order to explore the direction of causality between life satisfaction and political

participation one would need to work on panel data, with sufficient observations of

unemployed and even better unemployed youth.

Yet, there are a number of methodological challenges facing the study of unemployment

and political participation. When studying long-term unemployed youth through survey

methods in a context of low unemployment, it is challenging to accumulate a large enough

sample to run statistical models. Here I analyzed the contribution of unemployed youth

subjective well-being for political participation in contacting and protest activities. I found

that subjective well-being, in particular life satisfaction, played a different role for the

political participation of unemployed and employed youth. Although the small number of

respondents included in my sample made it difficult to reach conventional levels of sta-

tistical significance, the empirical results suggest that the effect was substantial.

In future research, life satisfaction should be used more systematically in the study of

political participation. In order to understand its contribution to different forms of political

participation, as already suggested by Flavin and Keane (2012), and to test its effects on

different sub-groups of the population. My research only started to untangle the complex

effects of life satisfaction working on unemployed and employed youth. Obviously, more

research is needed to understand the contribution of life satisfaction for the political

participation of unemployed persons, not focusing on youth.

Moreover, future studies should test the effect of life satisfaction on the political par-

ticipation of youth living in different socio-political contexts, in particular in contexts of

higher youth unemployment, to see whether the effects of life satisfaction are contingent

upon the case studied here. As mentioned in the introduction, youth unemployment is very

high in some European countries. In these contexts my findings suggest that unemployed

youth are not visible because they are too dissatisfied with their lives to engage in protest

activities. Considering the limited protest activities by unemployed youth in contexts of

high youth unemployment and the findings I presented here, we could hypothesize that life

satisfaction does not depend on unemployment rates. In fact, Oesch and Lipps (2012) show

that unemployment does not hurt less when there is more around. Moreover, in southern

Europe, familialist welfare regimes offer few or no social benefits to unemployed youth.

Thus, where youth unemployment is high, social protection is limited. Therefore, we may

find very few unemployed youth who are satisfied with their lives and this may explain the

absence of protest by unemployed youth in Europe. Future research should compare the

effects of unemployment on life satisfaction and protest activities across countries to test

this hypothesis. Alternatively, it should address first the effect of unemployment on

grievances and on life satisfaction in different contexts and, in a second step, test how

grievances and life satisfaction contribute to political participation.

Lastly, different measures of life satisfaction could be used and tested against specific

measures of deprivation that could account for the situation of unemployed youth. In

addition, the ideal design to test the impact of unemployment on political participation

through subjective well-being would require longitudinal data that would permit testing the

subjective well-being of unemployed youth before and after their lived experience of

unemployment, as well as their political participation at the time of unemployment.

Acknowledgments This paper was written while I was visiting scholar at the WissenschaftszentrumBerlin fur Sozialforschung. I am grateful to Professor Heike Solga for her hospitality and to Marco Giugni,Dietlind Stolle, Christian Lahusen, Lucio Bacarro, Jonas Pontusson, and two anonymous reviewers for

Subjective Well-Being and Political Participation

123

valuable comments. The academic stay at the Wissenschaftszentrum fur Sozialforschung was funded by theSwiss National Fund under the doc.mobility program. Results presented in this paper have been obtainedwithin the project ‘‘Youth, Unemployment, and Exclusion in Europe: A Multidimensional Approach toUnderstanding the Conditions and Prospects for Social and Political Integration of Young Unemployed’’(YOUNEX). This project was funded by the European Commission under the 7th Framework Programme(grant agreement no. 216122).

Appendix 1

See Table 5.

Appendix 2

See Table 6.

Table 5 Descriptive statistics on dependent and independent variables

Unemployed youth Employed youth Coding

Mean SD Mean SD Min Max

Dependent variables

Contacting activities 0.16 0.37 0.13 0.34 0 1

Protest activities 0.16 0.37 0.14 0.35 0 1

Independent variables

Main independent variable

Life satisfaction 4.09 2.00 5.19 1.51 0 7

Socio-economic status

Female 0.51 0.50 0.49 0.50 0 1

Age 28.26 4.44 28.83 4.26 18 35

Citizen of the country 0.50 0.50 0.61 0.49 0 1

Below secondary education 0.24 0.43 0.11 0.31 0 1

Secondary education 0.55 0.50 0.54 0.50 0 1

Tertiary education 0.21 0.41 0.35 0.48 0 1

Perceived financial difficulties 0.37 0.48 0.07 0.26 0 1

Civic attitudes

Political interest 0.42 0.50 0.43 0.50 0 1

Internal political efficacy 0.35 0.48 0.28 0.45 0 1

External political efficacy 0.61 0.49 0.54 0.50 0 1

Left self-placement 0.33 0.47 0.36 0.48 0 1

Right self-placement 0.21 0.41 0.35 0.48 0 1

Center or no self-placement 0.46 0.50 0.29 0.45 0 1

Social inclusion

Lives with partner 0.37 0.48 0.46 0.50 0 1

Lives with children 0.27 0.45 0.27 0.44 0 1

Contacts with friends 1.70 0.57 1.80 0.47 0 2

Associational membership 0.45 0.50 0.39 0.49 0 1

Health

Mental health (GHQ score on 4 items) 8.49 2.32 9.80 1.95 0 12

Observations 276 294

J. Lorenzini

123

Ta

ble

6L

ogis

tic

regre

ssio

ns

on

pro

test

acti

vit

ies

(exponen

tiat

edco

effi

cien

ts,

stan

dar

der

ror)

Mo

del

1M

od

el2

Mo

del

3M

od

el4

Mo

del

5M

odel

6M

odel

7

Lif

esa

tisf

acti

on

1.0

66

(0.4

93

)0

.48

9(0

.36

3)

0.4

19

(0.3

23

)0

.36

2(0

.29

6)

0.4

81

(0.3

69)

0.3

57

(0.2

77)

0.3

10

(0.2

70)

Lo

ng-t

erm

un

emp

loy

edy

ou

th1

.214

(0.2

97

)0

.52

6(0

.35

9)

0.4

37

(0.3

14

)0

.40

1(0

.29

6)

0.4

23

(0.2

99)

0.5

52

(0.3

79)

0.2

43

?(0

.19

2)

Inte

ract

ion

Un

emp

loy

ed*

Lif

esa

tisf

acti

on

3.4

05

(3.2

11)

5.3

43

?(5

.26

0)

5.2

75

(5.4

06)

4.2

96

(4.2

05)

3.5

08

(3.3

37)

11

.07

*(1

2.0

2)

Soci

o-e

conom

icst

atu

s

Educa

tion

(Ref

.bel

ow

seco

ndar

y)

Sec

ond

ary

1.8

58

(0.7

56

)2

.18

0?

(0.9

72)

Ter

tiar

y3

.465

**

(1.5

86

)3

.27

6*

(1.6

96)

Per

ceiv

edfi

nan

cial

dif

ficu

ltie

s1

.393

(0.4

43

)1

.30

5(0

.45

9)

Fem

ale

0.5

76

*(0

.14

1)

0.5

80

*(0

.15

9)

Ag

e0

.474

(0.2

45

)0

.17

1*

*(0

.10

4)

Cit

izen

of

the

cou

ntr

y1

.572

?(0

.39

9)

1.2

26

(0.3

45)

Civ

ica

ttit

udes

Po

liti

cal

inte

rest

2.6

34

**

*(0

.69

1)

2.5

07

**

(0.7

11)

Lef

t–ri

gh

tsc

ale

(Ref

.ce

nte

r)

Lef

tse

lfp

lace

men

t3

.37

3*

**

(1.0

13)

3.1

55

**

*(1

.02

4)

Rig

ht

self

pla

cem

ent

0.6

00

(0.2

45)

0.5

73

(0.2

44)

Inte

rnal

po

liti

cal

effi

cacy

0.8

89

(0.2

43)

0.8

48

(0.2

52)

Ex

tern

alp

oli

tica

lef

fica

cy0

.94

3(0

.24

1)

0.9

23

(0.2

52)

So

cia

lca

pit

al

Liv

esw

ith

par

tner

0.5

91

?(0

.17

0)

0.6

12

(0.1

99)

Liv

esw

ith

chil

dre

n1

.110

(0.3

43)

1.2

94

(0.4

65)

Co

nta

cts

wit

hfr

ien

ds

(Ref

.n

oco

nta

cts)

Few

con

tact

s1

.920

(1.5

71)

1.4

01

(1.2

54)

Fre

quen

tco

nta

cts

2.2

26

(1.7

12)

0.9

22

(0.7

89)

Ass

oci

atio

nal

mem

ber

ship

3.4

54

**

*(0

.86

9)

3.1

71

**

*(0

.88

0)

Subjective Well-Being and Political Participation

123

Ta

ble

6co

nti

nued

Mo

del

1M

od

el2

Mo

del

3M

od

el4

Mo

del

5M

odel

6M

odel

7

Hea

lth

Men

tal

hea

lth

(GH

Q)

2.9

63

(2.1

89)

v2(d

f_m

)0

.637

(2)

2.2

98

(3)

19

.59

(9)*

57

.57

(8)*

**

32

.04

(8)*

**

4.5

01

(4)

90

.64

(19

)**

*

AIC

48

9.0

48

9.3

48

4.0

44

4.1

46

9.6

48

9.1

43

3.0

N5

70

57

05

70

57

05

70

57

05

70

?p\

0.1

0;

*p\

0.0

5;

**

p\

0.0

1;

**

*p\

0.0

01

J. Lorenzini

123

Tab

le7

Logis

tic

regre

ssio

ns

on

conta

ctin

gac

tivit

ies

(exponen

tiat

edco

effi

cien

ts,

stan

dar

der

ror)

Mo

del

1M

odel

2M

od

el3

Mo

del

4M

odel

5M

od

el6

Mo

del

7

Lif

esa

tisf

acti

on

0.4

39

?(0

.19

9)

0.1

42

**

(0.1

03

)0

.14

1*

*(0

.10

7)

0.1

25

**

(0.0

935

)0

.131

**

(0.0

979

)0

.11

4*

*(0

.08

67

)0

.125

*(0

.10

2)

Lo

ng

-ter

mu

nem

plo

yed

yo

uth

1.1

12

(0.2

78)

0.3

50

?(0

.22

2)

0.2

66

?(0

.18

1)

0.3

54

(0.2

33)

0.3

07

?(0

.20

1)

0.3

65

(0.2

33)

0.2

85

?(0

.20

3)

Inte

ract

ion

Un

emp

loy

ed*

Lif

esa

tisf

acti

on

6.0

40

?(5

.55

8)

9.8

83

*(9

.62

0)

6.3

53

?(6

.03

5)

6.6

51

*(6

.27

5)

6.0

93

?(5

.63

9)

8.4

32

*(8

.59

3)

So

cio

-eco

no

mic

sta

tus

Educa

tion

(Ref

.bel

ow

seco

ndar

y)

Sec

ond

ary

1.3

91

(0.5

32)

1.7

23

(0.7

46

)

Ter

tiar

y1.4

83

(0.6

42)

1.6

02

(0.7

90)

Per

ceiv

edfi

nan

cial

dif

ficu

ltie

s1

.44

0(0

.44

7)

1.2

68

(0.4

35

)

Fem

ale

0.3

80

**

*(0

.09

88

)0

.453

**

(0.1

26

)

Ag

e4

.33

4*

*(2

.40

6)

2.9

73

?(1

.85

3)

Cit

izen

of

the

cou

ntr

y1

.00

2(0

.25

3)

0.8

78

(0.2

41

)

Civ

ica

ttit

udes

Po

liti

cal

inte

rest

2.3

66

**

*(0

.61

2)

1.9

95

*(0

.56

0)

Lef

t–ri

ght

scal

e(R

ef.

cente

r)

Lef

tse

lfp

lace

men

t1

.646

(0.5

24)

1.4

10

(0.4

86

)

Rig

ht

self

pla

cem

ent

1.9

72

*(0

.64

7)

2.0

87

*(0

.72

7)

Ap

pen

dix

3

See

Tab

le7.

Subjective Well-Being and Political Participation

123

Tab

le7

con

tin

ued

Mo

del

1M

odel

2M

od

el3

Mo

del

4M

odel

5M

od

el6

Mo

del

7

Inte

rnal

po

liti

cal

effi

cacy

1.6

51

*(0

.41

7)

1.6

50

?(0

.46

2)

Ex

tern

alp

oli

tica

lef

fica

cy1

.249

(0.3

19

)1

.141

(0.3

12

)

So

cia

lca

pit

al

Liv

esw

ith

par

tner

0.8

47

(0.2

42)

0.8

06

(0.2

50

)

Liv

esw

ith

chil

dre

n1

.219

(0.3

70)

1.2

35

(0.4

24

)

Co

nta

cts

wit

hfr

ien

ds

(Ref

.n

oco

nta

cts)

Few

con

tact

s0

.249

*(0

.15

1)

0.1

97

*(0

.13

1)

Fre

quen

tco

nta

cts

0.5

01

(0.2

50)

0.3

97

(0.2

32

)

Ass

oci

atio

nal

mem

ber

ship

3.1

21

**

*(0

.79

4)

2.9

26

**

*(0

.80

0)

Hea

lth

Men

tal

hea

lth

(GH

Q)

2.0

96

(1.5

49)

v2

(df_

m)

4.3

14

(2)

8.1

03

(3)*

34

.32

(9)*

**

35

.45

(8)*

**

36

.70

(8)*

**

9.1

19

(4)

77

.20

(19

)**

*

AIC

47

8.3

47

6.6

46

2.3

45

9.2

45

8.0

47

7.5

43

9.5

N5

70

57

05

70

57

05

70

57

05

70

?p\

0.1

0;

*p\

0.0

5;

**

p\

0.0

1;

**

*p\

0.0

01

J. Lorenzini

123

References

Almond, G. A., & Verba, S. (1963). The civic culture: Political attitudes and democracy in five nations.Princeton, NJ: Princeton University Press.

Banks, M. H., & Ullah, P. (1987). Political attitudes and voting among unemployed and employed youth.Journal of Adolescence, 10(2), 201–216.

Barnes, S. H., Farah, B. G., & Heunks, F. (1979). Personal Dissatisfaction. In S. H. Barnes & M. Kaase(Eds.), Political action: Mass participation in five western democracies (pp. 381–408). Beverly Hills,CA: Sage.

Bay, A.-H., & Blekesaune, M. (2002). Youth, unemployment and political marginalisation. InternationalJournal of Social Welfare, 11(2), 132–139.

Boes, S., & Winkelmann, R. (2010). The effect of income on general life satisfaction and dissatisfaction.Social Indicators Research, 95(1), 111–128. doi:10.1007/s11205-009-9452-7.

Breakwell, G. M. (1986). Political and attributional responses of the young short-term unemployed. PoliticalPsychology, 7(3), 575–586.

Burden, B., & Wichowsky, A. (2012). Economic hardship and voter turnout. Chicago: Paper presented atthe MidWest Political Science Association.

Bynner, J., & Ashford, S. (1994). Politics and participation: Some antecedents of young people’s attitudes tothe political system and political activity. European Journal of Social Psychology, 24(2), 223–236.doi:10.1002/ejsp.2420240202.

Cameron, A. C., & Trivedi, P. K. (2009). Microeconometrics using stata. College Station, TX: Stata Press.Chabanet, D. (2007). Chomage et exclusion sociale: l’echec europeen. Politique europEenne, 21(1),

157–187.Chabanet, D. (2008). When the unemployed challenge the European Union: The European marches as a

mode of externalization of protest. Mobilization, 13, 311–322.Clark, A. W. (1985). The effects of unemployment on political attitude. Journal of Sociology, 21(1),

100–108.Clark, A. E., & Oswald, A. J. (1994). Unhappiness and unemployment. The Economic Journal, 104(424),

648–659.Conway, M. M. (2000). Political participation in the United States (3rd ed.). Washington, DC: CQ Press.Dalton, R. J. (Ed.). (2006). Citizen politics: Public opinion and political parties in advanced industrial

democracies (4th ed.). Washington, DC: CQ Press.della Porta, D. (2008). Protest on unemployment: forms and opportunities. Mobilization, 13(3), 277–295.Dolan, P., Peasgood, T., & White, M. (2008). Do we really know what makes us happy? A review of the

economic literature on the factors associated with subjective well-being. Journal of Economic Psy-chology, 29(1), 94–122.

Dorn, D., Fischer, J., Kirchgassner, G., & Sousa-Poza, A. (2008). Direct democracy and life satisfactionrevisited: New evidence for Switzerland. Journal of Happiness Studies, 9(2), 227–255. doi:10.1007/s10902-007-9050-9.

Ervasti, H., & Venetoklis, T. (2010). Unemployment and subjective well-being. Acta Sociologica, 53(2),119–139. doi:10.1177/0001699310365624.

Faniel, J. (2004). Chomeurs en Belgique et en France: des mobilisations differentes. Revue internationale depolitique comparEe, 11(4), 493–506. doi:10.3917/ripc.114.0493.

Flavin, P., & Keane, M. (2012). Life satisfaction and political participation: Evidence from the UnitedStates. Journal of Happiness Studies, 1–16, doi:10.1007/s10902-011-9250-1.

Frey, B., & Stutzer, A. (2000). Happiness, economy and institutions. The Economic Journal, 110(466),918–938. doi:10.1111/1468-0297.00570.

Frey, B., & Stutzer, A. (2010). Happiness and public choice. Public Choice, 144(3), 557–573. doi:10.1007/s11127-010-9681-y.

Gallie, D., Paugam, S., & Jacobs, S. (2003). Unemployment, poverty, and social isolation: Is there a viciouscircle of social exclusion? European Societies, 5(1), 1–32. doi:10.1080/1461669032000057668.

Gaskell, G., & Smith, P. (1985). An investigation of youths’ attributions for unemployment and theirpolitical attitudes. Journal of Economic Psychology, 6(1), 65–80. doi:10.1016/0167-4870(85)90006-6.

Giugni, M. (2008). Welfare states, political opportunities, and the mobilization of the unemployed: A cross-national analysis. Mobilization, 13, 297–310.

Hammer, T. (2000). Mental health and social exclusion among unemployed youth in Scandinavia. Acomparative study. International Journal of Social Welfare, 9(1), 53–63. doi:10.1111/1468-2397.00108.

Jahoda, M., Lazarsfeld, P. F., & Zeisel, H. (2002 [1932]). Marienthal: The sociography of an unemployedcommunity New Brunswick, USA: Transaction Publishers.

Subjective Well-Being and Political Participation

123

Kriesi, H. (2008). Political mobilisation, political participation and the power of the vote. West EuropeanPolitics, 31(1), 147–168.

Kriesi, H., Koopmans, R., Duyvendak, J. W., & Giugni, M. G. (1995). New social movements in WesternEurope. Minneapolis: University of Minnesota Press.

Leighley, J. E. (1995). Attitudes, opportunities and incentives: A field essay on political participation.Political Research Quarterly, 48(1), 181–209.

Lorenzini, J. (2013). Unemployment and citizenship: Social and political participation of unemployed youthin Geneva. Universite de Geneve.

Marsh, A. (1977). Protest and political consciousness (Vol. 49, Sage library of social research). BeverlyHills, CA: Sage Publications.

Maurer, S. (2001). Les chomeurs en action (decembre 1997–mars 1998): Mobilisation collective et res-sources compensatoires (Logiques sociales). Paris: L’Harmattan.

Mood, C. (2010). Logistic regression: Why we cannot do what we think we can do, and what we can doabout it. European Sociological Review, 26(1), 67–82. doi:10.1093/esr/jcp006.

Oesch, D., & Lipps, O. (2012). Does unemployment hurt less if there is more of it around? A panel analysisof life satisfaction in Germany and Switzerland. European Sociological Review (first published online).doi:10.1093/esr/jcs071.

Paugam, S. (2006). L’epreuve du chomage: une rupture cumulative des liens sociaux? Revue europeennedes sciences sociales, XLIV–135.

Rosenstone, S. J. (1982). Economic adversity and voter turnout. American Journal of Political Science,26(1), 25–46.

Rosenstone, S. J., & Hansen, J. M. (1993). Mobilization, participation, and democracy in America (newtopics in politics). New York: Macmillan.

Schlozman, K. L., & Verba, S. (1979). Injury to insult: Unemployment, class, and political response.Cambridge: Harvard University Press.

Scott, W. J., & Acock, A. C. (1979). Socioeconomic status, unemployment experience, and political par-ticipation: A disentangling of main and interaction effects. Political Behavior, 1(4), 361–381.

Stutzer, A., & Lalive, R. (2004). The role of social work norms in job searching and subjective well-being.Journal of the European Economic Association, 2(4), 696–719. doi:10.1162/1542476041423331.

Tavits, M. (2008). Representation, corruption, and subjective well-being. Comparative Political Studies,41(12), 1607–1630. doi:10.1177/0010414007308537.

Teorell, J., Torcal, M., & Montero, J. R. (2007). Political participation: Mapping the terrain. In J. W. vanDeth, J. R. Montero, & A. Westholm (Eds.), Citizenship and involvement in European democracies: Acomparative analysis (pp. 334–357). London: Routledge.

Thoits, P. A., & Hewitt, L. N. (2001). Volunteer work and well-being. Journal of Health and SocialBehavior, 42(2), 115–131.

Veenhoven, R. (1988). The utility of happiness. Social Indicators Research, 20(4), 333–354. doi:10.1007/bf00302332.

Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism in Americanpolitics. Cambridge, MA: Harvard University Press.

Weitz-Shapiro, R., & Winters, M. S. (2011). The link between voting and life satisfaction in Latin America.Latin American Politics and Society, 53(4), 101–126. doi:10.1111/j.1548-2456.2011.00135.x.

Whelan, C., & McGinnity, F. (2000). Unemployment and satisfaction: A European analysis. In D. Gallie &S. Paugam (Eds.), Welfare regimes and the experience of unemployment in Europe (Repr ed.). Oxford:Oxford University Press.

Winefield, A. H. (1997). Introduction to the psychological effects of youth unemployment: Internationalperspectives. Journal of Adolescence, 20, 237–241.

Winkelmann, L., & Winkelmann, R. (1998). Why are the unemployed so unhappy? Evidence from paneldata. Economica, 65(257), 1–15. doi:10.1111/1468-0335.00111.

J. Lorenzini

123